SlideShare ist ein Scribd-Unternehmen logo
1 von 24
Downloaden Sie, um offline zu lesen
Background Pre-computation Simulation Study Conclusion
Pre-computation for ABC in image analysis
Matt Moores1,2 Kerrie Mengersen1,2 Christian Robert3,4
1Mathematical Sciences School, Queensland University of Technology,
Brisbane, Australia
2Institute for Health and Biomedical Innovation, QUT Kelvin Grove
3CEREMADE, Universit´e Paris Dauphine, France
4CREST, INSEE, France
MCMSki IV, Chamonix 2014
Background Pre-computation Simulation Study Conclusion
Outline
1 Background
Approximate Bayesian Computation (ABC)
Sequential Monte Carlo (SMC-ABC)
Hidden Potts model
2 Pre-computation
3 Simulation Study
Background Pre-computation Simulation Study Conclusion
Background
Image analysis often involves:
Large datasets, with millions of pixels
Multiple images with similar characteristics
For example: satellite remote sensing (Landsat), computed
tomography (CT)
Table : Scale of common types of images
Number Landsat CT slices
of pixels (90m2/px) (512×512)
26 0.06km2
. . .
56 14.06km2
0.1
106 900.00km2
3.8
156 10251.56km2
43.5
Background Pre-computation Simulation Study Conclusion
Approximate Bayesian Computation (ABC)
Algorithm 1 ABC rejection sampler
1: for all iterations t ∈ 1 . . . T do
2: Draw independent proposal θ ∼ π(θ)
3: Generate x ∼ f(·|θ )
4: if |ρ(x) − ρ(y)| < then
5: set θt ← θ
6: else
7: set θt ← θt−1
8: end if
9: end for
Pritchard, Seielstad, Perez-Lezaun & Feldman (1999) Mol. Biol. Evol. 16(12)
Marin, Pudlo, Robert & Ryder (2012) Stat. Comput. 22(6)
Background Pre-computation Simulation Study Conclusion
Adaptive ABC using Sequential Monte Carlo (SMC-ABC)
Algorithm 2 SMC-ABC
1: Draw N particles θi ∼ π(θ)
2: Generate pseudo-data xi,m ∼ f(·|θi)
3: repeat
4: Adaptively select ABC tolerance t
5: Update importance weights ωi for each particle
6: if effective sample size (ESS) < Nmin then
7: Resample particles according to their weights
8: end if
9: Update particles using random walk proposal
(with adaptive RWMH bandwidth σ2
t )
10: until
naccept
N < 0.015 or t = 0
Del Moral, Doucet, & Jasra (2012) Stat. Comput. 22(5)
Liu (2001) Monte Carlo Strategies in Scientific Computing New York: Springer
Background Pre-computation Simulation Study Conclusion
Motivation
Computational cost is dominated by simulation of pseudo-data
e.g. Hidden Potts model in image analysis
(Grelaud et al. 2009, Everitt 2012)
Model fitting with ABC can be separated into:
Learning about the summary statistic, given the parameter
ρ(x) | θ
Choosing parameter values, given a summary statistic
θ | ρ(y)
For latent models, an additional step of learning about the
summary statistic, given the data: ρ(z) | y, θ
Grelaud, Robert, Marin, Rodolphe & Taly (2009) Bayesian Analysis 4(2)
Everitt (2012) JCGS 21(4)
Background Pre-computation Simulation Study Conclusion
hidden Markov random field
Joint distribution of observed pixel intensities yi ∈ y
and latent labels zi ∈ z:
Pr(y, z|µ, σ2
) ∝ L(y|µ, σ2
, z)π(µ|σ2
)π(σ2
)π(z|β)π(β) (1)
Additive Gaussian noise:
yi|zi =j
iid
∼ N µj, σ2
j (2)
Potts model:
π(zi|zi∼ , β) =
exp {β i∼ δ(zi, z )}
k
j=1 exp {β i∼ δ(j, z )}
(3)
Potts (1952) Proceedings of the Cambridge Philosophical Society 48(1)
Background Pre-computation Simulation Study Conclusion
Inverse Temperature
Background Pre-computation Simulation Study Conclusion
Doubly-intractable likelihood
p(β|z) = C(β)−1
π(β) exp {β S(z)} (4)
The normalising constant of the Potts model has computational
complexity of O(n2kn), since it involves a sum over all possible
combinations of the labels z ∈ Z:
C(β) =
z∈Z
exp {β S(z)} (5)
S(z) is the sufficient statistic of the Potts model:
S(z) =
i∼ ∈L
δ(zi, z ) (6)
where L is the set of all unique neighbour pairs.
Background Pre-computation Simulation Study Conclusion
Pre-computation
The distribution of ρ(x) | θ is independent of the data
By simulating pseudo-data for values of θ, we can create a
mapping function ˆf(θ) to approximate E[ρ(x)|θ]
This mapping function can be reused across multiple datasets,
amortising its computational cost
By mapping directly from θ → ρ(x), we avoid the need to simulate
pseudo-data during model fitting
Background Pre-computation Simulation Study Conclusion
Sufficient statistic of the Potts model
0.0 0.5 1.0 1.5 2.0 2.5 3.0
1000015000200002500030000
β
S(z)
(a) E(S(z)|β)
0.0 0.5 1.0 1.5 2.0 2.5 3.0050100150200250
β
σ(S(z))
(b) σ(S(z)|β)
Figure : Distribution of S(z) | β for n = 56
, k = 3
Background Pre-computation Simulation Study Conclusion
Scalable SMC-ABC for the hidden Potts model
Algorithm 3 SMC-ABC using precomputed ˆf(β)
1: Draw N particles βi ∼ π0(β)
2: Approximate sufficient statistics S(xi,m) ≈ ˆf(βi)
3: repeat
4: Update S(zt)|y, πt(β)
5: Adaptively select ABC tolerance t
6: Update importance weights ωi for each particle
7: if effective sample size (ESS) < Nmin then
8: Resample particles according to their weights
9: end if
10: Update particles using random walk proposal
(with adaptive RWMH bandwidth σ2
t )
11: until
naccept
N < 0.015 or t < 10−9 or t ≥ 100
Background Pre-computation Simulation Study Conclusion
Simulation Study
20 images, n = 125 × 125, k = 3:
β ∼ U(0, 1.005)
z ∼ f(·|β) using 2000 iterations of Swendsen-Wang
µj ∼ N 0, 1002
1
σ2
j
∼ Γ (1, 100)
Comparison of 2 ABC algorithms:
Scalable SMC-ABC using precomputed ˆf(β)
Standard SMC-ABC using 500 iterations of Gibbs sampling
Swendsen & Wang (1987) Physical Review Letters 58
Background Pre-computation Simulation Study Conclusion
Accuracy of posterior estimates for β
0.2 0.4 0.6 0.8 1.0
0.00.20.40.60.81.0
β
posteriordistribution
(a) pseudo-data
0.2 0.4 0.6 0.8 1.0
0.00.20.40.60.81.0
β
posteriordistribution
(b) pre-computed
Background Pre-computation Simulation Study Conclusion
Distribution of posterior sampling error for β
algorithm
error
0.0
0.2
0.4
0.6
Pseudo−data Pre−computed
Background Pre-computation Simulation Study Conclusion
Improvement in runtime
Pseudo−data Pre−computed
0.51.02.05.010.020.050.0100.0
algorithm
elapsedtime(hours)
(a) elapsed (wall clock) time
Pseudo−data Pre−computed
51020501002005001000
algorithm
CPUtime(hours)
(b) CPU time
Background Pre-computation Simulation Study Conclusion
Summary
Scalability of SMC-ABC can be improved by pre-computing an
approximate mapping θ → ρ(x)
Pre-computation took 8 minutes on a 16 core Xeon server
Average runtime for SMC-ABC improved from 74.4 hours to
39 minutes
The mapping function represents the nonlinear, heteroskedastic
relationship between the parameter and the summary statistic.
This method could be extended to multivariate applications, such
as estimating both β and k for the hidden Potts model.
Appendix
Acknowledgements
I gratefully acknowledge the financial support received from:
Mathematical Sciences School,
Queensland University of Technology, Brisbane, Australia
Institute for Health and Biomedical Innovation, QUT
Bayesian section of the American Statistical Association
International Society for Bayesian Analysis
BayesComp section of ISBA
CEREMADE, Universit´e Paris Dauphine, France
Department of Economics, University of Warwick, UK
Computational resources and services used in this work were
provided by the HPC and Research Support Group, QUT.
Appendix
For Further Reading I
Jun S. Liu
Monte Carlo Strategies in Scientific Computing
Springer-Verlag, 2001.
Pierre Del Moral, Arnaud Doucet & Ajay Jasra
An adaptive sequential Monte Carlo method for approximate Bayesian
computation.
Statistics & Computing, 22(5): 1009–20, 2012.
Richard Everitt
Bayesian Parameter Estimation for Latent Markov Random Fields and
Social Networks.
J. Comput. Graph. Stat., 21(4): 940–60, 2012.
A. Grelaud, C. P. Robert, J.-M. Marin, F. Rodolphe & J.-F. Taly
ABC likelihood-free methods for model choice in Gibbs random fields.
Bayesian Analysis, 4(2): 317–36, 2009.
Appendix
For Further Reading II
J.-M. Marin, P. Pudlo, C. P. Robert & R. J. Ryder
Approximate Bayesian computational methods.
Statistics & Computing, 22(6): 1167–80, 2012.
Renfrey B. Potts
Some generalized order-disorder transformations.
Proc. Cambridge Philosophical Society, 48(1): 106–9, 1952.
J. K. Pritchard, M. T. Seielstad, A. Perez-Lezaun & M. W. Feldman
Population growth of human Y chromosomes: a study of Y chromosome
microsatellites
Mol. Biol. Evol., 16(12): 1791–8, 1999.
R. H. Swendsen & J.-S. Wang
Nonuniversal critical dynamics in Monte Carlo simulations.
Physical Review Letters, 58: 86–8, 1987.
Appendix
ABC tolerance levels
Iteration
0 20 40 60 80 100
0200040006000
ε
σ
(a)
Iteration
0 20 40 60 80 100
0200040006000800010000
ε
σ
(b)
Appendix
Sufficient Statistic
Iteration
S(z)
0 20 40 60 80 100
15000155001600016500170001750018000
(a)
Iteration
S(z)
0 20 40 60 80 100
100001200014000160001800020000
(b)
Appendix
RWMH acceptance rate
Iteration
proportionaccepted
0 20 40 60 80 100
0.00.20.40.60.81.0
(a)
Iteration
proportionaccepted
0 20 40 60 80 100
0.00.20.40.60.81.0
(b)
Appendix
Effective Sample Size
Iteration
ESS
0 20 40 60 80 100
0200040006000800010000
(a)
Iteration
ESS
0 20 40 60 80 100
0200040006000800010000
(b)

Weitere ähnliche Inhalte

Was ist angesagt?

ABC short course: model choice chapter
ABC short course: model choice chapterABC short course: model choice chapter
ABC short course: model choice chapterChristian Robert
 
ABC short course: introduction chapters
ABC short course: introduction chaptersABC short course: introduction chapters
ABC short course: introduction chaptersChristian Robert
 
Laplace's Demon: seminar #1
Laplace's Demon: seminar #1Laplace's Demon: seminar #1
Laplace's Demon: seminar #1Christian Robert
 
Macrocanonical models for texture synthesis
Macrocanonical models for texture synthesisMacrocanonical models for texture synthesis
Macrocanonical models for texture synthesisValentin De Bortoli
 
NCE, GANs & VAEs (and maybe BAC)
NCE, GANs & VAEs (and maybe BAC)NCE, GANs & VAEs (and maybe BAC)
NCE, GANs & VAEs (and maybe BAC)Christian Robert
 
Optimal interval clustering: Application to Bregman clustering and statistica...
Optimal interval clustering: Application to Bregman clustering and statistica...Optimal interval clustering: Application to Bregman clustering and statistica...
Optimal interval clustering: Application to Bregman clustering and statistica...Frank Nielsen
 
A nonlinear approximation of the Bayesian Update formula
A nonlinear approximation of the Bayesian Update formulaA nonlinear approximation of the Bayesian Update formula
A nonlinear approximation of the Bayesian Update formulaAlexander Litvinenko
 
better together? statistical learning in models made of modules
better together? statistical learning in models made of modulesbetter together? statistical learning in models made of modules
better together? statistical learning in models made of modulesChristian Robert
 
Pattern learning and recognition on statistical manifolds: An information-geo...
Pattern learning and recognition on statistical manifolds: An information-geo...Pattern learning and recognition on statistical manifolds: An information-geo...
Pattern learning and recognition on statistical manifolds: An information-geo...Frank Nielsen
 
ABC short course: survey chapter
ABC short course: survey chapterABC short course: survey chapter
ABC short course: survey chapterChristian Robert
 

Was ist angesagt? (20)

ABC workshop: 17w5025
ABC workshop: 17w5025ABC workshop: 17w5025
ABC workshop: 17w5025
 
ABC-Gibbs
ABC-GibbsABC-Gibbs
ABC-Gibbs
 
ABC-Gibbs
ABC-GibbsABC-Gibbs
ABC-Gibbs
 
QMC Program: Trends and Advances in Monte Carlo Sampling Algorithms Workshop,...
QMC Program: Trends and Advances in Monte Carlo Sampling Algorithms Workshop,...QMC Program: Trends and Advances in Monte Carlo Sampling Algorithms Workshop,...
QMC Program: Trends and Advances in Monte Carlo Sampling Algorithms Workshop,...
 
ABC short course: model choice chapter
ABC short course: model choice chapterABC short course: model choice chapter
ABC short course: model choice chapter
 
ABC short course: introduction chapters
ABC short course: introduction chaptersABC short course: introduction chapters
ABC short course: introduction chapters
 
Laplace's Demon: seminar #1
Laplace's Demon: seminar #1Laplace's Demon: seminar #1
Laplace's Demon: seminar #1
 
Gtti 10032021
Gtti 10032021Gtti 10032021
Gtti 10032021
 
Macrocanonical models for texture synthesis
Macrocanonical models for texture synthesisMacrocanonical models for texture synthesis
Macrocanonical models for texture synthesis
 
Poster(3)-1
Poster(3)-1Poster(3)-1
Poster(3)-1
 
NCE, GANs & VAEs (and maybe BAC)
NCE, GANs & VAEs (and maybe BAC)NCE, GANs & VAEs (and maybe BAC)
NCE, GANs & VAEs (and maybe BAC)
 
Optimal interval clustering: Application to Bregman clustering and statistica...
Optimal interval clustering: Application to Bregman clustering and statistica...Optimal interval clustering: Application to Bregman clustering and statistica...
Optimal interval clustering: Application to Bregman clustering and statistica...
 
Program on Mathematical and Statistical Methods for Climate and the Earth Sys...
Program on Mathematical and Statistical Methods for Climate and the Earth Sys...Program on Mathematical and Statistical Methods for Climate and the Earth Sys...
Program on Mathematical and Statistical Methods for Climate and the Earth Sys...
 
A nonlinear approximation of the Bayesian Update formula
A nonlinear approximation of the Bayesian Update formulaA nonlinear approximation of the Bayesian Update formula
A nonlinear approximation of the Bayesian Update formula
 
better together? statistical learning in models made of modules
better together? statistical learning in models made of modulesbetter together? statistical learning in models made of modules
better together? statistical learning in models made of modules
 
Pattern learning and recognition on statistical manifolds: An information-geo...
Pattern learning and recognition on statistical manifolds: An information-geo...Pattern learning and recognition on statistical manifolds: An information-geo...
Pattern learning and recognition on statistical manifolds: An information-geo...
 
Program on Quasi-Monte Carlo and High-Dimensional Sampling Methods for Applie...
Program on Quasi-Monte Carlo and High-Dimensional Sampling Methods for Applie...Program on Quasi-Monte Carlo and High-Dimensional Sampling Methods for Applie...
Program on Quasi-Monte Carlo and High-Dimensional Sampling Methods for Applie...
 
Lausanne 2019 #2
Lausanne 2019 #2Lausanne 2019 #2
Lausanne 2019 #2
 
ABC short course: survey chapter
ABC short course: survey chapterABC short course: survey chapter
ABC short course: survey chapter
 
11 clusadvanced
11 clusadvanced11 clusadvanced
11 clusadvanced
 

Ähnlich wie Pre-computation for ABC in image analysis

R package bayesImageS: Scalable Inference for Intractable Likelihoods
R package bayesImageS: Scalable Inference for Intractable LikelihoodsR package bayesImageS: Scalable Inference for Intractable Likelihoods
R package bayesImageS: Scalable Inference for Intractable LikelihoodsMatt Moores
 
Workshop in honour of Don Poskitt and Gael Martin
Workshop in honour of Don Poskitt and Gael MartinWorkshop in honour of Don Poskitt and Gael Martin
Workshop in honour of Don Poskitt and Gael MartinChristian Robert
 
Unbiased Bayes for Big Data
Unbiased Bayes for Big DataUnbiased Bayes for Big Data
Unbiased Bayes for Big DataChristian Robert
 
Interpolation of-geofield-parameters
Interpolation of-geofield-parametersInterpolation of-geofield-parameters
Interpolation of-geofield-parametersCemal Ardil
 
bayesImageS: Bayesian computation for medical Image Segmentation using a hidd...
bayesImageS: Bayesian computation for medical Image Segmentation using a hidd...bayesImageS: Bayesian computation for medical Image Segmentation using a hidd...
bayesImageS: Bayesian computation for medical Image Segmentation using a hidd...Matt Moores
 
Solving inverse problems via non-linear Bayesian Update of PCE coefficients
Solving inverse problems via non-linear Bayesian Update of PCE coefficientsSolving inverse problems via non-linear Bayesian Update of PCE coefficients
Solving inverse problems via non-linear Bayesian Update of PCE coefficientsAlexander Litvinenko
 
Patch Matching with Polynomial Exponential Families and Projective Divergences
Patch Matching with Polynomial Exponential Families and Projective DivergencesPatch Matching with Polynomial Exponential Families and Projective Divergences
Patch Matching with Polynomial Exponential Families and Projective DivergencesFrank Nielsen
 
"An adaptive modular approach to the mining of sensor network ...
"An adaptive modular approach to the mining of sensor network ..."An adaptive modular approach to the mining of sensor network ...
"An adaptive modular approach to the mining of sensor network ...butest
 
CARI-2020, Application of LSTM architectures for next frame forecasting in Se...
CARI-2020, Application of LSTM architectures for next frame forecasting in Se...CARI-2020, Application of LSTM architectures for next frame forecasting in Se...
CARI-2020, Application of LSTM architectures for next frame forecasting in Se...Mokhtar SELLAMI
 
Intro to Model Selection
Intro to Model SelectionIntro to Model Selection
Intro to Model Selectionchenhm
 
Likelihood free computational statistics
Likelihood free computational statisticsLikelihood free computational statistics
Likelihood free computational statisticsPierre Pudlo
 
Stratified sampling and resampling for approximate Bayesian computation
Stratified sampling and resampling for approximate Bayesian computationStratified sampling and resampling for approximate Bayesian computation
Stratified sampling and resampling for approximate Bayesian computationUmberto Picchini
 
Investigation on the Pattern Synthesis of Subarray Weights for Low EMI Applic...
Investigation on the Pattern Synthesis of Subarray Weights for Low EMI Applic...Investigation on the Pattern Synthesis of Subarray Weights for Low EMI Applic...
Investigation on the Pattern Synthesis of Subarray Weights for Low EMI Applic...IOSRJECE
 
Data-Driven Motion Estimation With Spatial Adaptation
Data-Driven Motion Estimation With Spatial AdaptationData-Driven Motion Estimation With Spatial Adaptation
Data-Driven Motion Estimation With Spatial AdaptationCSCJournals
 

Ähnlich wie Pre-computation for ABC in image analysis (20)

R package bayesImageS: Scalable Inference for Intractable Likelihoods
R package bayesImageS: Scalable Inference for Intractable LikelihoodsR package bayesImageS: Scalable Inference for Intractable Likelihoods
R package bayesImageS: Scalable Inference for Intractable Likelihoods
 
Workshop in honour of Don Poskitt and Gael Martin
Workshop in honour of Don Poskitt and Gael MartinWorkshop in honour of Don Poskitt and Gael Martin
Workshop in honour of Don Poskitt and Gael Martin
 
Unbiased Bayes for Big Data
Unbiased Bayes for Big DataUnbiased Bayes for Big Data
Unbiased Bayes for Big Data
 
Interpolation of-geofield-parameters
Interpolation of-geofield-parametersInterpolation of-geofield-parameters
Interpolation of-geofield-parameters
 
bayesImageS: Bayesian computation for medical Image Segmentation using a hidd...
bayesImageS: Bayesian computation for medical Image Segmentation using a hidd...bayesImageS: Bayesian computation for medical Image Segmentation using a hidd...
bayesImageS: Bayesian computation for medical Image Segmentation using a hidd...
 
Litvinenko nlbu2016
Litvinenko nlbu2016Litvinenko nlbu2016
Litvinenko nlbu2016
 
Solving inverse problems via non-linear Bayesian Update of PCE coefficients
Solving inverse problems via non-linear Bayesian Update of PCE coefficientsSolving inverse problems via non-linear Bayesian Update of PCE coefficients
Solving inverse problems via non-linear Bayesian Update of PCE coefficients
 
Patch Matching with Polynomial Exponential Families and Projective Divergences
Patch Matching with Polynomial Exponential Families and Projective DivergencesPatch Matching with Polynomial Exponential Families and Projective Divergences
Patch Matching with Polynomial Exponential Families and Projective Divergences
 
Dycops2019
Dycops2019 Dycops2019
Dycops2019
 
"An adaptive modular approach to the mining of sensor network ...
"An adaptive modular approach to the mining of sensor network ..."An adaptive modular approach to the mining of sensor network ...
"An adaptive modular approach to the mining of sensor network ...
 
CARI-2020, Application of LSTM architectures for next frame forecasting in Se...
CARI-2020, Application of LSTM architectures for next frame forecasting in Se...CARI-2020, Application of LSTM architectures for next frame forecasting in Se...
CARI-2020, Application of LSTM architectures for next frame forecasting in Se...
 
Automatic bayesian cubature
Automatic bayesian cubatureAutomatic bayesian cubature
Automatic bayesian cubature
 
Intro to Model Selection
Intro to Model SelectionIntro to Model Selection
Intro to Model Selection
 
Likelihood free computational statistics
Likelihood free computational statisticsLikelihood free computational statistics
Likelihood free computational statistics
 
Stratified sampling and resampling for approximate Bayesian computation
Stratified sampling and resampling for approximate Bayesian computationStratified sampling and resampling for approximate Bayesian computation
Stratified sampling and resampling for approximate Bayesian computation
 
Investigation on the Pattern Synthesis of Subarray Weights for Low EMI Applic...
Investigation on the Pattern Synthesis of Subarray Weights for Low EMI Applic...Investigation on the Pattern Synthesis of Subarray Weights for Low EMI Applic...
Investigation on the Pattern Synthesis of Subarray Weights for Low EMI Applic...
 
the ABC of ABC
the ABC of ABCthe ABC of ABC
the ABC of ABC
 
Data-Driven Motion Estimation With Spatial Adaptation
Data-Driven Motion Estimation With Spatial AdaptationData-Driven Motion Estimation With Spatial Adaptation
Data-Driven Motion Estimation With Spatial Adaptation
 
asymptotics of ABC
asymptotics of ABCasymptotics of ABC
asymptotics of ABC
 
2009 asilomar
2009 asilomar2009 asilomar
2009 asilomar
 

Mehr von Matt Moores

Bayesian Inference and Uncertainty Quantification for Inverse Problems
Bayesian Inference and Uncertainty Quantification for Inverse ProblemsBayesian Inference and Uncertainty Quantification for Inverse Problems
Bayesian Inference and Uncertainty Quantification for Inverse ProblemsMatt Moores
 
bayesImageS: an R package for Bayesian image analysis
bayesImageS: an R package for Bayesian image analysisbayesImageS: an R package for Bayesian image analysis
bayesImageS: an R package for Bayesian image analysisMatt Moores
 
Exploratory Analysis of Multivariate Data
Exploratory Analysis of Multivariate DataExploratory Analysis of Multivariate Data
Exploratory Analysis of Multivariate DataMatt Moores
 
Importing satellite imagery into R from NASA and the U.S. Geological Survey
Importing satellite imagery into R from NASA and the U.S. Geological SurveyImporting satellite imagery into R from NASA and the U.S. Geological Survey
Importing satellite imagery into R from NASA and the U.S. Geological SurveyMatt Moores
 
Accelerating Pseudo-Marginal MCMC using Gaussian Processes
Accelerating Pseudo-Marginal MCMC using Gaussian ProcessesAccelerating Pseudo-Marginal MCMC using Gaussian Processes
Accelerating Pseudo-Marginal MCMC using Gaussian ProcessesMatt Moores
 
R package 'bayesImageS': a case study in Bayesian computation using Rcpp and ...
R package 'bayesImageS': a case study in Bayesian computation using Rcpp and ...R package 'bayesImageS': a case study in Bayesian computation using Rcpp and ...
R package 'bayesImageS': a case study in Bayesian computation using Rcpp and ...Matt Moores
 
Variational Bayes
Variational BayesVariational Bayes
Variational BayesMatt Moores
 
Informative Priors for Segmentation of Medical Images
Informative Priors for Segmentation of Medical ImagesInformative Priors for Segmentation of Medical Images
Informative Priors for Segmentation of Medical ImagesMatt Moores
 

Mehr von Matt Moores (9)

Bayesian Inference and Uncertainty Quantification for Inverse Problems
Bayesian Inference and Uncertainty Quantification for Inverse ProblemsBayesian Inference and Uncertainty Quantification for Inverse Problems
Bayesian Inference and Uncertainty Quantification for Inverse Problems
 
bayesImageS: an R package for Bayesian image analysis
bayesImageS: an R package for Bayesian image analysisbayesImageS: an R package for Bayesian image analysis
bayesImageS: an R package for Bayesian image analysis
 
Exploratory Analysis of Multivariate Data
Exploratory Analysis of Multivariate DataExploratory Analysis of Multivariate Data
Exploratory Analysis of Multivariate Data
 
Importing satellite imagery into R from NASA and the U.S. Geological Survey
Importing satellite imagery into R from NASA and the U.S. Geological SurveyImporting satellite imagery into R from NASA and the U.S. Geological Survey
Importing satellite imagery into R from NASA and the U.S. Geological Survey
 
Accelerating Pseudo-Marginal MCMC using Gaussian Processes
Accelerating Pseudo-Marginal MCMC using Gaussian ProcessesAccelerating Pseudo-Marginal MCMC using Gaussian Processes
Accelerating Pseudo-Marginal MCMC using Gaussian Processes
 
R package 'bayesImageS': a case study in Bayesian computation using Rcpp and ...
R package 'bayesImageS': a case study in Bayesian computation using Rcpp and ...R package 'bayesImageS': a case study in Bayesian computation using Rcpp and ...
R package 'bayesImageS': a case study in Bayesian computation using Rcpp and ...
 
Variational Bayes
Variational BayesVariational Bayes
Variational Bayes
 
Parallel R
Parallel RParallel R
Parallel R
 
Informative Priors for Segmentation of Medical Images
Informative Priors for Segmentation of Medical ImagesInformative Priors for Segmentation of Medical Images
Informative Priors for Segmentation of Medical Images
 

Kürzlich hochgeladen

Software Quality Assurance Interview Questions
Software Quality Assurance Interview QuestionsSoftware Quality Assurance Interview Questions
Software Quality Assurance Interview QuestionsArshad QA
 
Short Story: Unveiling the Reasoning Abilities of Large Language Models by Ke...
Short Story: Unveiling the Reasoning Abilities of Large Language Models by Ke...Short Story: Unveiling the Reasoning Abilities of Large Language Models by Ke...
Short Story: Unveiling the Reasoning Abilities of Large Language Models by Ke...kellynguyen01
 
Right Money Management App For Your Financial Goals
Right Money Management App For Your Financial GoalsRight Money Management App For Your Financial Goals
Right Money Management App For Your Financial GoalsJhone kinadey
 
Learn the Fundamentals of XCUITest Framework_ A Beginner's Guide.pdf
Learn the Fundamentals of XCUITest Framework_ A Beginner's Guide.pdfLearn the Fundamentals of XCUITest Framework_ A Beginner's Guide.pdf
Learn the Fundamentals of XCUITest Framework_ A Beginner's Guide.pdfkalichargn70th171
 
HR Software Buyers Guide in 2024 - HRSoftware.com
HR Software Buyers Guide in 2024 - HRSoftware.comHR Software Buyers Guide in 2024 - HRSoftware.com
HR Software Buyers Guide in 2024 - HRSoftware.comFatema Valibhai
 
The Ultimate Test Automation Guide_ Best Practices and Tips.pdf
The Ultimate Test Automation Guide_ Best Practices and Tips.pdfThe Ultimate Test Automation Guide_ Best Practices and Tips.pdf
The Ultimate Test Automation Guide_ Best Practices and Tips.pdfkalichargn70th171
 
Optimizing AI for immediate response in Smart CCTV
Optimizing AI for immediate response in Smart CCTVOptimizing AI for immediate response in Smart CCTV
Optimizing AI for immediate response in Smart CCTVshikhaohhpro
 
Steps To Getting Up And Running Quickly With MyTimeClock Employee Scheduling ...
Steps To Getting Up And Running Quickly With MyTimeClock Employee Scheduling ...Steps To Getting Up And Running Quickly With MyTimeClock Employee Scheduling ...
Steps To Getting Up And Running Quickly With MyTimeClock Employee Scheduling ...MyIntelliSource, Inc.
 
How To Use Server-Side Rendering with Nuxt.js
How To Use Server-Side Rendering with Nuxt.jsHow To Use Server-Side Rendering with Nuxt.js
How To Use Server-Side Rendering with Nuxt.jsAndolasoft Inc
 
Shapes for Sharing between Graph Data Spaces - and Epistemic Querying of RDF-...
Shapes for Sharing between Graph Data Spaces - and Epistemic Querying of RDF-...Shapes for Sharing between Graph Data Spaces - and Epistemic Querying of RDF-...
Shapes for Sharing between Graph Data Spaces - and Epistemic Querying of RDF-...Steffen Staab
 
Unlocking the Future of AI Agents with Large Language Models
Unlocking the Future of AI Agents with Large Language ModelsUnlocking the Future of AI Agents with Large Language Models
Unlocking the Future of AI Agents with Large Language Modelsaagamshah0812
 
5 Signs You Need a Fashion PLM Software.pdf
5 Signs You Need a Fashion PLM Software.pdf5 Signs You Need a Fashion PLM Software.pdf
5 Signs You Need a Fashion PLM Software.pdfWave PLM
 
+971565801893>>SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHAB...
+971565801893>>SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHAB...+971565801893>>SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHAB...
+971565801893>>SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHAB...Health
 
W01_panagenda_Navigating-the-Future-with-The-Hitchhikers-Guide-to-Notes-and-D...
W01_panagenda_Navigating-the-Future-with-The-Hitchhikers-Guide-to-Notes-and-D...W01_panagenda_Navigating-the-Future-with-The-Hitchhikers-Guide-to-Notes-and-D...
W01_panagenda_Navigating-the-Future-with-The-Hitchhikers-Guide-to-Notes-and-D...panagenda
 
The Real-World Challenges of Medical Device Cybersecurity- Mitigating Vulnera...
The Real-World Challenges of Medical Device Cybersecurity- Mitigating Vulnera...The Real-World Challenges of Medical Device Cybersecurity- Mitigating Vulnera...
The Real-World Challenges of Medical Device Cybersecurity- Mitigating Vulnera...ICS
 
Try MyIntelliAccount Cloud Accounting Software As A Service Solution Risk Fre...
Try MyIntelliAccount Cloud Accounting Software As A Service Solution Risk Fre...Try MyIntelliAccount Cloud Accounting Software As A Service Solution Risk Fre...
Try MyIntelliAccount Cloud Accounting Software As A Service Solution Risk Fre...MyIntelliSource, Inc.
 
A Secure and Reliable Document Management System is Essential.docx
A Secure and Reliable Document Management System is Essential.docxA Secure and Reliable Document Management System is Essential.docx
A Secure and Reliable Document Management System is Essential.docxComplianceQuest1
 

Kürzlich hochgeladen (20)

CHEAP Call Girls in Pushp Vihar (-DELHI )🔝 9953056974🔝(=)/CALL GIRLS SERVICE
CHEAP Call Girls in Pushp Vihar (-DELHI )🔝 9953056974🔝(=)/CALL GIRLS SERVICECHEAP Call Girls in Pushp Vihar (-DELHI )🔝 9953056974🔝(=)/CALL GIRLS SERVICE
CHEAP Call Girls in Pushp Vihar (-DELHI )🔝 9953056974🔝(=)/CALL GIRLS SERVICE
 
Software Quality Assurance Interview Questions
Software Quality Assurance Interview QuestionsSoftware Quality Assurance Interview Questions
Software Quality Assurance Interview Questions
 
Short Story: Unveiling the Reasoning Abilities of Large Language Models by Ke...
Short Story: Unveiling the Reasoning Abilities of Large Language Models by Ke...Short Story: Unveiling the Reasoning Abilities of Large Language Models by Ke...
Short Story: Unveiling the Reasoning Abilities of Large Language Models by Ke...
 
Right Money Management App For Your Financial Goals
Right Money Management App For Your Financial GoalsRight Money Management App For Your Financial Goals
Right Money Management App For Your Financial Goals
 
Learn the Fundamentals of XCUITest Framework_ A Beginner's Guide.pdf
Learn the Fundamentals of XCUITest Framework_ A Beginner's Guide.pdfLearn the Fundamentals of XCUITest Framework_ A Beginner's Guide.pdf
Learn the Fundamentals of XCUITest Framework_ A Beginner's Guide.pdf
 
HR Software Buyers Guide in 2024 - HRSoftware.com
HR Software Buyers Guide in 2024 - HRSoftware.comHR Software Buyers Guide in 2024 - HRSoftware.com
HR Software Buyers Guide in 2024 - HRSoftware.com
 
The Ultimate Test Automation Guide_ Best Practices and Tips.pdf
The Ultimate Test Automation Guide_ Best Practices and Tips.pdfThe Ultimate Test Automation Guide_ Best Practices and Tips.pdf
The Ultimate Test Automation Guide_ Best Practices and Tips.pdf
 
Optimizing AI for immediate response in Smart CCTV
Optimizing AI for immediate response in Smart CCTVOptimizing AI for immediate response in Smart CCTV
Optimizing AI for immediate response in Smart CCTV
 
Steps To Getting Up And Running Quickly With MyTimeClock Employee Scheduling ...
Steps To Getting Up And Running Quickly With MyTimeClock Employee Scheduling ...Steps To Getting Up And Running Quickly With MyTimeClock Employee Scheduling ...
Steps To Getting Up And Running Quickly With MyTimeClock Employee Scheduling ...
 
How To Use Server-Side Rendering with Nuxt.js
How To Use Server-Side Rendering with Nuxt.jsHow To Use Server-Side Rendering with Nuxt.js
How To Use Server-Side Rendering with Nuxt.js
 
Shapes for Sharing between Graph Data Spaces - and Epistemic Querying of RDF-...
Shapes for Sharing between Graph Data Spaces - and Epistemic Querying of RDF-...Shapes for Sharing between Graph Data Spaces - and Epistemic Querying of RDF-...
Shapes for Sharing between Graph Data Spaces - and Epistemic Querying of RDF-...
 
Vip Call Girls Noida ➡️ Delhi ➡️ 9999965857 No Advance 24HRS Live
Vip Call Girls Noida ➡️ Delhi ➡️ 9999965857 No Advance 24HRS LiveVip Call Girls Noida ➡️ Delhi ➡️ 9999965857 No Advance 24HRS Live
Vip Call Girls Noida ➡️ Delhi ➡️ 9999965857 No Advance 24HRS Live
 
Unlocking the Future of AI Agents with Large Language Models
Unlocking the Future of AI Agents with Large Language ModelsUnlocking the Future of AI Agents with Large Language Models
Unlocking the Future of AI Agents with Large Language Models
 
5 Signs You Need a Fashion PLM Software.pdf
5 Signs You Need a Fashion PLM Software.pdf5 Signs You Need a Fashion PLM Software.pdf
5 Signs You Need a Fashion PLM Software.pdf
 
+971565801893>>SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHAB...
+971565801893>>SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHAB...+971565801893>>SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHAB...
+971565801893>>SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHAB...
 
W01_panagenda_Navigating-the-Future-with-The-Hitchhikers-Guide-to-Notes-and-D...
W01_panagenda_Navigating-the-Future-with-The-Hitchhikers-Guide-to-Notes-and-D...W01_panagenda_Navigating-the-Future-with-The-Hitchhikers-Guide-to-Notes-and-D...
W01_panagenda_Navigating-the-Future-with-The-Hitchhikers-Guide-to-Notes-and-D...
 
The Real-World Challenges of Medical Device Cybersecurity- Mitigating Vulnera...
The Real-World Challenges of Medical Device Cybersecurity- Mitigating Vulnera...The Real-World Challenges of Medical Device Cybersecurity- Mitigating Vulnera...
The Real-World Challenges of Medical Device Cybersecurity- Mitigating Vulnera...
 
Try MyIntelliAccount Cloud Accounting Software As A Service Solution Risk Fre...
Try MyIntelliAccount Cloud Accounting Software As A Service Solution Risk Fre...Try MyIntelliAccount Cloud Accounting Software As A Service Solution Risk Fre...
Try MyIntelliAccount Cloud Accounting Software As A Service Solution Risk Fre...
 
A Secure and Reliable Document Management System is Essential.docx
A Secure and Reliable Document Management System is Essential.docxA Secure and Reliable Document Management System is Essential.docx
A Secure and Reliable Document Management System is Essential.docx
 
Microsoft AI Transformation Partner Playbook.pdf
Microsoft AI Transformation Partner Playbook.pdfMicrosoft AI Transformation Partner Playbook.pdf
Microsoft AI Transformation Partner Playbook.pdf
 

Pre-computation for ABC in image analysis

  • 1. Background Pre-computation Simulation Study Conclusion Pre-computation for ABC in image analysis Matt Moores1,2 Kerrie Mengersen1,2 Christian Robert3,4 1Mathematical Sciences School, Queensland University of Technology, Brisbane, Australia 2Institute for Health and Biomedical Innovation, QUT Kelvin Grove 3CEREMADE, Universit´e Paris Dauphine, France 4CREST, INSEE, France MCMSki IV, Chamonix 2014
  • 2. Background Pre-computation Simulation Study Conclusion Outline 1 Background Approximate Bayesian Computation (ABC) Sequential Monte Carlo (SMC-ABC) Hidden Potts model 2 Pre-computation 3 Simulation Study
  • 3. Background Pre-computation Simulation Study Conclusion Background Image analysis often involves: Large datasets, with millions of pixels Multiple images with similar characteristics For example: satellite remote sensing (Landsat), computed tomography (CT) Table : Scale of common types of images Number Landsat CT slices of pixels (90m2/px) (512×512) 26 0.06km2 . . . 56 14.06km2 0.1 106 900.00km2 3.8 156 10251.56km2 43.5
  • 4. Background Pre-computation Simulation Study Conclusion Approximate Bayesian Computation (ABC) Algorithm 1 ABC rejection sampler 1: for all iterations t ∈ 1 . . . T do 2: Draw independent proposal θ ∼ π(θ) 3: Generate x ∼ f(·|θ ) 4: if |ρ(x) − ρ(y)| < then 5: set θt ← θ 6: else 7: set θt ← θt−1 8: end if 9: end for Pritchard, Seielstad, Perez-Lezaun & Feldman (1999) Mol. Biol. Evol. 16(12) Marin, Pudlo, Robert & Ryder (2012) Stat. Comput. 22(6)
  • 5. Background Pre-computation Simulation Study Conclusion Adaptive ABC using Sequential Monte Carlo (SMC-ABC) Algorithm 2 SMC-ABC 1: Draw N particles θi ∼ π(θ) 2: Generate pseudo-data xi,m ∼ f(·|θi) 3: repeat 4: Adaptively select ABC tolerance t 5: Update importance weights ωi for each particle 6: if effective sample size (ESS) < Nmin then 7: Resample particles according to their weights 8: end if 9: Update particles using random walk proposal (with adaptive RWMH bandwidth σ2 t ) 10: until naccept N < 0.015 or t = 0 Del Moral, Doucet, & Jasra (2012) Stat. Comput. 22(5) Liu (2001) Monte Carlo Strategies in Scientific Computing New York: Springer
  • 6. Background Pre-computation Simulation Study Conclusion Motivation Computational cost is dominated by simulation of pseudo-data e.g. Hidden Potts model in image analysis (Grelaud et al. 2009, Everitt 2012) Model fitting with ABC can be separated into: Learning about the summary statistic, given the parameter ρ(x) | θ Choosing parameter values, given a summary statistic θ | ρ(y) For latent models, an additional step of learning about the summary statistic, given the data: ρ(z) | y, θ Grelaud, Robert, Marin, Rodolphe & Taly (2009) Bayesian Analysis 4(2) Everitt (2012) JCGS 21(4)
  • 7. Background Pre-computation Simulation Study Conclusion hidden Markov random field Joint distribution of observed pixel intensities yi ∈ y and latent labels zi ∈ z: Pr(y, z|µ, σ2 ) ∝ L(y|µ, σ2 , z)π(µ|σ2 )π(σ2 )π(z|β)π(β) (1) Additive Gaussian noise: yi|zi =j iid ∼ N µj, σ2 j (2) Potts model: π(zi|zi∼ , β) = exp {β i∼ δ(zi, z )} k j=1 exp {β i∼ δ(j, z )} (3) Potts (1952) Proceedings of the Cambridge Philosophical Society 48(1)
  • 8. Background Pre-computation Simulation Study Conclusion Inverse Temperature
  • 9. Background Pre-computation Simulation Study Conclusion Doubly-intractable likelihood p(β|z) = C(β)−1 π(β) exp {β S(z)} (4) The normalising constant of the Potts model has computational complexity of O(n2kn), since it involves a sum over all possible combinations of the labels z ∈ Z: C(β) = z∈Z exp {β S(z)} (5) S(z) is the sufficient statistic of the Potts model: S(z) = i∼ ∈L δ(zi, z ) (6) where L is the set of all unique neighbour pairs.
  • 10. Background Pre-computation Simulation Study Conclusion Pre-computation The distribution of ρ(x) | θ is independent of the data By simulating pseudo-data for values of θ, we can create a mapping function ˆf(θ) to approximate E[ρ(x)|θ] This mapping function can be reused across multiple datasets, amortising its computational cost By mapping directly from θ → ρ(x), we avoid the need to simulate pseudo-data during model fitting
  • 11. Background Pre-computation Simulation Study Conclusion Sufficient statistic of the Potts model 0.0 0.5 1.0 1.5 2.0 2.5 3.0 1000015000200002500030000 β S(z) (a) E(S(z)|β) 0.0 0.5 1.0 1.5 2.0 2.5 3.0050100150200250 β σ(S(z)) (b) σ(S(z)|β) Figure : Distribution of S(z) | β for n = 56 , k = 3
  • 12. Background Pre-computation Simulation Study Conclusion Scalable SMC-ABC for the hidden Potts model Algorithm 3 SMC-ABC using precomputed ˆf(β) 1: Draw N particles βi ∼ π0(β) 2: Approximate sufficient statistics S(xi,m) ≈ ˆf(βi) 3: repeat 4: Update S(zt)|y, πt(β) 5: Adaptively select ABC tolerance t 6: Update importance weights ωi for each particle 7: if effective sample size (ESS) < Nmin then 8: Resample particles according to their weights 9: end if 10: Update particles using random walk proposal (with adaptive RWMH bandwidth σ2 t ) 11: until naccept N < 0.015 or t < 10−9 or t ≥ 100
  • 13. Background Pre-computation Simulation Study Conclusion Simulation Study 20 images, n = 125 × 125, k = 3: β ∼ U(0, 1.005) z ∼ f(·|β) using 2000 iterations of Swendsen-Wang µj ∼ N 0, 1002 1 σ2 j ∼ Γ (1, 100) Comparison of 2 ABC algorithms: Scalable SMC-ABC using precomputed ˆf(β) Standard SMC-ABC using 500 iterations of Gibbs sampling Swendsen & Wang (1987) Physical Review Letters 58
  • 14. Background Pre-computation Simulation Study Conclusion Accuracy of posterior estimates for β 0.2 0.4 0.6 0.8 1.0 0.00.20.40.60.81.0 β posteriordistribution (a) pseudo-data 0.2 0.4 0.6 0.8 1.0 0.00.20.40.60.81.0 β posteriordistribution (b) pre-computed
  • 15. Background Pre-computation Simulation Study Conclusion Distribution of posterior sampling error for β algorithm error 0.0 0.2 0.4 0.6 Pseudo−data Pre−computed
  • 16. Background Pre-computation Simulation Study Conclusion Improvement in runtime Pseudo−data Pre−computed 0.51.02.05.010.020.050.0100.0 algorithm elapsedtime(hours) (a) elapsed (wall clock) time Pseudo−data Pre−computed 51020501002005001000 algorithm CPUtime(hours) (b) CPU time
  • 17. Background Pre-computation Simulation Study Conclusion Summary Scalability of SMC-ABC can be improved by pre-computing an approximate mapping θ → ρ(x) Pre-computation took 8 minutes on a 16 core Xeon server Average runtime for SMC-ABC improved from 74.4 hours to 39 minutes The mapping function represents the nonlinear, heteroskedastic relationship between the parameter and the summary statistic. This method could be extended to multivariate applications, such as estimating both β and k for the hidden Potts model.
  • 18. Appendix Acknowledgements I gratefully acknowledge the financial support received from: Mathematical Sciences School, Queensland University of Technology, Brisbane, Australia Institute for Health and Biomedical Innovation, QUT Bayesian section of the American Statistical Association International Society for Bayesian Analysis BayesComp section of ISBA CEREMADE, Universit´e Paris Dauphine, France Department of Economics, University of Warwick, UK Computational resources and services used in this work were provided by the HPC and Research Support Group, QUT.
  • 19. Appendix For Further Reading I Jun S. Liu Monte Carlo Strategies in Scientific Computing Springer-Verlag, 2001. Pierre Del Moral, Arnaud Doucet & Ajay Jasra An adaptive sequential Monte Carlo method for approximate Bayesian computation. Statistics & Computing, 22(5): 1009–20, 2012. Richard Everitt Bayesian Parameter Estimation for Latent Markov Random Fields and Social Networks. J. Comput. Graph. Stat., 21(4): 940–60, 2012. A. Grelaud, C. P. Robert, J.-M. Marin, F. Rodolphe & J.-F. Taly ABC likelihood-free methods for model choice in Gibbs random fields. Bayesian Analysis, 4(2): 317–36, 2009.
  • 20. Appendix For Further Reading II J.-M. Marin, P. Pudlo, C. P. Robert & R. J. Ryder Approximate Bayesian computational methods. Statistics & Computing, 22(6): 1167–80, 2012. Renfrey B. Potts Some generalized order-disorder transformations. Proc. Cambridge Philosophical Society, 48(1): 106–9, 1952. J. K. Pritchard, M. T. Seielstad, A. Perez-Lezaun & M. W. Feldman Population growth of human Y chromosomes: a study of Y chromosome microsatellites Mol. Biol. Evol., 16(12): 1791–8, 1999. R. H. Swendsen & J.-S. Wang Nonuniversal critical dynamics in Monte Carlo simulations. Physical Review Letters, 58: 86–8, 1987.
  • 21. Appendix ABC tolerance levels Iteration 0 20 40 60 80 100 0200040006000 ε σ (a) Iteration 0 20 40 60 80 100 0200040006000800010000 ε σ (b)
  • 22. Appendix Sufficient Statistic Iteration S(z) 0 20 40 60 80 100 15000155001600016500170001750018000 (a) Iteration S(z) 0 20 40 60 80 100 100001200014000160001800020000 (b)
  • 23. Appendix RWMH acceptance rate Iteration proportionaccepted 0 20 40 60 80 100 0.00.20.40.60.81.0 (a) Iteration proportionaccepted 0 20 40 60 80 100 0.00.20.40.60.81.0 (b)
  • 24. Appendix Effective Sample Size Iteration ESS 0 20 40 60 80 100 0200040006000800010000 (a) Iteration ESS 0 20 40 60 80 100 0200040006000800010000 (b)